Evaluation of a commercial real-time multiplex PCR assay for the detection of bovine mastitis pathogens directly from milk

Graduate Theses and Dissertations Iowa State University Capstones, Theses and Dissertations 2016 Evaluation of a commercial real-time multiplex PCR assay for the detection of bovine mastitis pathogens directly from milk Lacey Marshall Lund Iowa State University Follow this and additional works at: http://lib.dr.iastate.edu/etd Part of the Microbiology Commons, and the Veterinary Medicine Commons Recommended Citation Marshall Lund, Lacey, "Evaluation of a commercial real-time multiplex PCR assay for the detection of bovine mastitis pathogens directly from milk" (2016). Graduate Theses and Dissertations. 15765. http://lib.dr.iastate.edu/etd/15765 This Thesis is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact digirep@iastate.edu.

Evaluation of a commercial real-time multiplex PCR assay for the detection of bovine mastitis pathogens directly from milk by Lacey Marshall Lund A thesis submitted to the graduate faculty in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Major: Microbiology Program of Study Committee: Philip Gauger, Co-Major Professor Timothy Frana, Co-Major Professor Karen Harmon Leo Timms Iowa State University Ames, Iowa 2016 Copyright Lacey Marshall Lund, 2016. All rights reserved.

ii DEDICATION This thesis is dedicated to my parents Kathy and Greg who have blessed me with their guidance to this point over great distance with love and patience. Everything you have taught me has brought me to a place where I have found purpose, passion and endless opportunities. I will never forget how fortunate I am. This thesis is also dedicated to those whose greatest challenge in life is themselves.

iii TABLE OF CONTENTS Page NOMENCLATURE... ACKNOWLEDGMENTS... ABSTRACT... v vii ix CHAPTER 1 INTRODUCTION: THESIS FORMATTING... 1 CHAPTER 2 REVIEW OF BOVINE MASTITIS ETIOLOGY AND DIAGNOSTIC METHODS... 2 Introduction... 2 The U.S. dairy industry... 2 Etiology of mastitis... 3 Clinical mastitis... 4 Subclinical mastitis... 4 Immunity of the mammary gland... 5 Somatic cell count... 6 Vaccination... 7 Contagious and environmental mastitis... 8 Contagious mastitis... 8 Staphylococcus aureus... 8 Streptococcus agalactiae... 9 Mycoplasma species... 10 Environmental mastitis... 11 Environmental Streptococcus species... 11 Coliforms... 12 Enterococcus species... 13 Viruses... 14 Diagnosis of bovine mastitis... 14 PCR/Nucleic acid tests for the detection of bovine mastitis... 15 PathoProof Mastitis PCR... 18 Future use of PathoProof PCR... 23 References... 25

iv CHAPTER 3 QUANTITATIVE AND QUALITATIVE EVALUATION OF REAL-TIME PCR IN THE DETECTION OF BOVINE MASTITIS PATHOGENS... 33 Abstract... 33 Introduction... 34 Methods and materials... 37 Specificity evaluation... 37 Cultures.. 37 DNA extraction... 38 Real-time PCR... 38 Sensitivity... 39 Results... 39 Specificity... 39 Sensitivity... 38 Discussion... 41 Conclusion... 41 References... 46 Tables Table 1. Targets detected in each PCR reaction by PathoProof PCR 50 Table 2. Specificity data analyzed by SDS and Norden software. 50 Table 3. Specificity analysis of non-target bacterial isolates, MALDI-TOF MS scores and PathoProof qpcr data analysis using System Detection Software and Norden Mastitis Studio software... 51 Table 4. Analytical sensitivity analysis using S. aureus, S. dysgalactiae and E. coli diluted in PBS and milk.... 51 CHAPTER 4 EVALUATION OF CONVENTIONAL CULTURE AND REAL-TIME PCR IN THE DETECTION OF BOVINE MASTITIS AGENTS IN COWS BEING TREATED FOR MASTITIS... 52 Abstract... 52 Introduction... 53 Methods and materials... 56 Cow selection... 56 Bacterial culture... 57 Real-time PCR... 57 Results... 58 Bacterial culture... 59 Real-time PCR... 59 Discussion... 60 Conclusion... 62 References... 64 Tables... 69 Table 1. Targets detected in each PCR reaction by PathoProof PCR... 69 Table 2. List of cows and bacteria isolated from day 0 cultures identified using MALDI-TOF MS... 69

v Table 3. Culture results for cows on days 0, 3, 7, 14 and 30... 70 Table 4. PCR results for cows on days 0, 3, 7, 14 and 30 with average Cts 70 Figures 71 Figure 1. Percent qpcr and bacterial culture positive cows at 0, 3, 7, 14 and 30 days post treatment.. 71 CHAPTER 5 GENERAL CONCLUSIONS... 72 APPENDIX ADDITIONAL COW DATA.. 74 Table 1. Individual cow data including target, culture and PCR results for days 0, 3, 7, 14 and 30.. 74

vi NOMENCLATURE ADSA ATCC BC BTM BTSCC CFU CNS Ct DMSCC ESCC FDA IAC IMI LOD LPS MALDI-TOF MS American Dairy Science Association American type culture collection Bacterial culture Bulk tank milk Bulk tank somatic cell count Colony forming unit Coagulase-negative Staphylococcus Cycle threshold Direct microscopy somatic cell count Electronic somatic cell count Food and Drug Administration Internal amplification control Intramammary infection Limit of detection Lipopolysaccharide Matrix-assisted laser desorption/ionization-time of flight mass spectrometry ml N/A NMC PAMP PBS Milliliter Non applicable National Mastitis Council Pathogen associated molecular patterns Phosphate-buffered saline

vii PCR SCC SDS TBC TLR qpcr USDA Polymerase Chain Reaction Somatic cell count Sequence detection software Total bacteria count Toll-like receptor Quantitative Polymerase Chain Reaction United States Department of Agriculture

viii ACKNOWLEDGMENTS I would like to take the opportunity to acknowledge the support I have received since beginning my graduate studies. I would like to first thank Dr. Timothy Frana for his patience and guidance in all aspects of my graduate education and for giving me opportunity to present data at multiple conferences. Secondly I would like to express gratitude to Dr. Philip Gauger for his willingness to assume leadership of my committee last May. Dr. Gauger s positive attitude make him excellent to work with. I have benefitted from his expertise in diagnostic microbiology in ways that have shown me new directions to pursue in the presentation of this data. Dr. Karen Harmon provided much support with real-time PCR technology and interpretation. I have learned a great deal about real-time PCR from Dr. Harmon when I lacked experience with molecular diagnostics. Dr. Harmon s expertise was very helpful during much of the trouble shooting using this PCR assay. I would like to thank Dr. Leo Timms for his expertise in the area of bovine mastitis from which I benefitted during my undergraduate and graduate careers. Dr. Timms is more than a mentor to his students by being a friend and an example of what it is to serve others. I also received support from the Bacteriology Section in the Veterinary Diagnostic Laboratory who have become family over the past 5 years. I have spent much of my time in the Clinical Microbiology Laboratory where I could be part of a team that has accomplished so much for the VDL. My lab mates in Clinical Microbiology have been there to assist in my research, lending their time and skill while delaying their own responsibilities. Lab mates

ix Alissa Koons, Aric McDaniel, David Frisk, laboratory supervisor Danielle Kenne, Marissa Gregory and Hallie Warneke who have become close friends as well as colleagues. I feel very fortunate to have received training, and advising from Joann Kinyon during her time at Iowa State. Joann s remarkable patience and her vast knowledge of microbiology make it a privilege to work with her. She has become a friend and a role model to me. Doctors Pat Gorden and Mike Kleinhenz of the FSVM department gave their time and knowledge to help develop this study. Dr. Kleinhenz was very helpful in collecting milk samples for me when I was not able to. I would like to acknowledge the ISU Dairy Farm for use to the animals and cow data for this study. Mary Healey and Joe Detrick helped facilitate this. Alex and Kevin Blood also allowed me to use their cows and cow data for this study. Both Alex and his staff members went out of their way to help with collecting samples and are a pleasure to work with. My family and friends have been a strong source of emotional support. My mother Kathy, my father Greg and my brothers Andrew and Treye have helped make the distance seem less distant with their unending support. My wonderful friends Taysha, John and Jessica who have taught me so much and enriched my time at Iowa State. Much gratitude to GDZ who has steadied me in the most uncertain times. Without the encouragement of these people I would not have been able to accomplish this degree.

x ABSTRACT Bovine mastitis, prevalent in dairy cattle, is often caused by a bacterial infection in the mammary gland. Bovine mastitis is costly to the dairy industry for loss of milk production. Causative bacteria are determined by microbiological culture. Culture remains popular for its low cost, simple procedure and interpretation however, has limitations. Culture based assays are subjective, timely and may not support fastidious organisms. Nearly 30% of clinical mastitis cases are culture-negative especially in cows treated with antibiotics. Treated cows with negative milk cultures may indeed still be infected, shedding low numbers of bacteria that do not appear in culture causing false negative results. One potential alternative to culture is a commercial assay, PathoProof Mastitis PCR Assay (ThermoFisher Scientific, Waltham, MA). PathoProof is capable of detecting 11 mastitis-causing bacteria from milk. To evaluate PathoProof, the analytical specificity and limit of detection (LOD) was determined using 20 culture isolates of target and non-target bacteria. The LOD was determined by inoculating phosphate-buffered saline and milk with three different bacteria. Serial dilutions, standard plate count and PCR were performed. Further evaluation used cows that were treated with antibiotics for mastitis. Milk samples were collected from cows on days 0, 3, 7, 14 and 30 post-treatment. Samples were evaluated using culture and PathoProof PCR. PathoProof PCR only detected target bacteria from a group of 20 target and nontarget isolates resulting in an analytical specificity of 100%. Average LOD ranged from 10 3 to 10 4 CFU/mL and 10 1 to 10 3 CFU per PCR reaction, relatively high values compared to

xi previous investigations of other mastitis PCR assays. The high LOD suggests concerns about false negative results from cows shedding bacteria at low levels. Data from 25 cows treated for mastitis were used to compare culture to PathoProof. More cows were PCR-positive on days 3, 7, 14 and 30 post-treatment demonstrating that PathoProof may be helpful in detecting bacteria in milk from treated cows. Information from PathoProof may be useful in evaluating efficacy of treatment and assist veterinarians and producers in making decisions. Further investigation into the assay s sensitivity and quantitative abilities is needed to better determine its value.

1 CHAPTER I INTRODUCTION: THESIS FORMATTING This thesis is organized into 4 chapters. Chapter 1 outlines the format of the thesis. Chapter 2 entitled Review of bovine mastitis etiology and diagnostic methods reviews general information about mastitis and current research in mastitis diagnostics. Chapter 3, Quantitative and qualitative evaluation of real-time PCR in the detection of bovine mastitis pathogens follows chapter 2. Chapter 4 is titled, Evaluation of conventional bacteriological culture and real-time PCR for detection of bacteria in milk from cows treated for mastitis followed by chapter 5 General Conclusions. Tables and figures adjoining each chapter will be found at the end of the chapter after the list of references.

2 CHAPTER 2. REVIEW OF BOVINE MASTITIS ETIOLOGY AND DIAGNOSTIC METHODS Introduction Mastitis or intramammary infection refers to inflammation of the mammary gland and may affect all mammals. It has particular importance in the dairy industry affecting the quantity and quality of milk produced by infected cows resulting in significant economic losses. There are nearly 9.3 million dairy cows in the United States responsible for 100 million tons of milk (USDA-ERS, 2015) produced yearly. Purported as the perfect food, milk and milk products represent a significant source of protein and minerals required in the human diet. Maximizing milk production while maintaining the welfare of cows is a complex system of balancing animal nutrition, housing, animal reproduction, milking procedures and health. Bovine mastitis is the most common cause of decreased milk production in dairy cattle which further implies the importance of prevention and control of this costly disease and protection of the global food supply (ADSA and Jones, 2006). The ever increasing need for resources to support the growing global population places the burden on food producers to maximize efficiency to meet increased demands. The U.S. dairy industry The U.S. dairy industry has experienced dramatic changes due to rapid growth during the past century. The American Dairy Science Association (ADSA) was formed to increase dairy production through advancements in animal health, genetic selection and farm

3 technologies. Since their establishment in 1906, the ADSA has collected data regarding the status of the dairy industry in the U.S along with government agencies such as the U.S. Department of Agriculture (USDA) and the Food and Drug Administration (FDA). The previous century s dairy industry was considerably smaller and less efficient when dairy cattle remained in numbers less than ten to provide families with a source of food rather than a source of income. Between 1930 and 2014, the annual per capita consumption of dairy products in the U.S. increased to 614 pounds (USDA-ERS, 2015). Due to increased demands, the average size of dairy herds increased from 5 to 115 cows with an average annual milk production of 4,500 pounds in 1930 to nearly 20,000 pounds per cow in 2006 (ADSA and Jones, 2006). This advancement could not have come without great strides in farming technologies. Automated milking became standard in the 1940s by use of vacuum milking units that attached to all 4 teats to collect milk through controlled pulsation (ADSA and Jones, 2006). Vacuum pulsation is used in modern milking facilities and reduces milking time and labor per cow. Automatic milking has also increased milk yield through proper emptying of the udder and has helped limit the transfer of microorganisms (Thompson, 1980). Etiology of bovine mastitis Since the advent of automated milking and the efficiency it brings, dairy operations have been able to expand herd size and operations by decreasing the time, labor and materials spent milking each cow. These advancements in technology have also improved sanitation methods but even in well-sanitized dairy operations, cows are still constantly exposed to microorganisms making mastitis a common threat.

4 Clinical mastitis Clinical mastitis is the presence of inflammation of the mammary gland and may be attributed to injury or infection from any pathogenic microorganism due to invasion through the teat canal. Clinical signs are most often detected during premilking procedures that are performed by farm employees. Premilking procedures include the application of an antimicrobial solution to each teat, removal of excess soil and manure with cloth towels and the removal of foremilk also known as forestripping prior to attaching milking equipment (Hogan et al., 1999). In between milking times, bacteria present on the skin or environmental organisms from soil or manure may migrate into the teat canal and possibly further into the mammary gland. To prevent these organisms from entering the bulk tank and potential crosscontamination between cows, stripped foremilk is observed for color and consistency. Visible changes in the milk s consistency may include clotting, flaking and the presence of blood. Indications of clinical mastitis are not limited to changes in the milk. External redness and swelling of the udder, tenderness to the touch, heat and palpable abscesses may also be present in clinical mastitis cases (Hogan et al., 1999). Subclinical mastitis Subclinical mastitis is defined as the absence of physical signs of inflammation to the udder and is the most common form of mastitis (Orlandini, 2011). It is primarily detected by increased somatic cell count (SCC) in the milk when increased numbers of white blood cells are recruited to the mammary gland due to mild infection or trauma (Orlandini, 2011). The SCC is expressed in cells per milliliter (ml) of milk and maintains high significance in the dairy industry and is routinely measured on individual cows and from milk collected from the

5 bulk tank. Somatic cell counts of 200,000 cells per ml or higher is indicative of inflammation (Hogan et al., 1999). Immunity of the mammary gland The mammary gland is protected by physical and chemical factors that impede the invasion of microorganisms. Bacteria must enter through the teat orifice that is controlled by sphincter muscles that relax during milk removal or let-down and requires 2 hours to fully contract after completion of milking (Srivastava et al., 2015). If microorganisms pass through the teat orifice they contact the teat canal, a cylindrical duct located inside the teat opening. The stratified epithelium lining the teat canal secretes keratin, a waxy, physical barrier that forms a plug that exits the teat during initial milk ejection and reforms after milking is completed (Rainard and Riollet, 2006). Keratin also provides a chemical defense with bacteriostatic and bactericidal compounds that include long chain fatty acids. A previous study demonstrated an infection rate of approximately 26.3% for quarters that had the keratin removed prior to milking followed by exposure to Streptococcus agalactiae after the milking process suggesting the importance of keratin as a preventative against mastitis (Capuco et al., 1992). Despite keratin s protective effect, some bacteria have the ability to migrate into the mammary gland and cause infection. Innate immune responses are critical during the initial stage of mastitis to respond to infection by different pathogens after they invade the mammary gland. The first to recognize the release of bacterial compounds are toll-like receptors (TLR), proteins embedded in leukocytes and epithelial cell membranes that recognize pathogen associated molecular patterns (PAMPS) such as lipopolysaccharides or peptidoglycan fragments (Rainard and Riollet, 2006; Salyers and Whitt, 2002; Wellnitz and Bruckmaier, 2012). After the

6 stimulation of TLR s by invading bacteria inflammation is mediated by polymorphonuclear neutrophilic (PMN) leukocytes which are recruited in vast numbers from the blood to the mammary tissue (Wellnitz and Bruckmaier, 2012). PMNs engulf and destroy the bacteria using intracellular granules that contain bactericidal enzymes and proteins such as superoxide ions, hypochlorite, hydrogen peroxide and hydrolytic enzymes (Wellnitz and Bruckmaier, 2012). When expended, PMNs will lyse and are subsequently engulfed by macrophages and destroyed (Salyers and Whitt, 2002). Somatic Cell Count Leukocytes inevitably end up in milk due to their massive numbers in response to infection in the mammary gland. In bovine mastitis leukocytes, macrophages, secretory and squamous cells are collectively referred to as somatic cells that are measured in individual and bulk tank samples, a value called somatic cell count (SCC) (Norman et al., 2011). An SCC level reflects the health of the udder or the health of the herd. SCC also has significance in quality determination by the Food and Drug Administration, the governing body of milk regulation. According to the 2015 Grade A Pasteurized Milk Ordinance (PMO), milk of grade A status may not have a bulk tank somatic cell count greater than 750,000 cells/ml of raw milk (FDA, 2015). Although somatic cells in milk have not been found to pose a health risk to humans (Hogan, 2005), SCC is a reflection of the hygiene practices of the operation (Moxley et al., 1978). For the determination of mastitis, limits have been set at 200,000 cells/ml or greater as an indicator of inflammation (Norman et al., 2011). SCC is particularly useful in diagnosing subclinical mastitis as there are no visible clinical signs in the milk and has become a routine practice in mastitis diagnostics. Methods to determine SCC as approved by

7 the FDA are direct microscopic somatic cell count (DMSCC) or electronic somatic cell count (ESCC). DMSCC is the reference method performed by dispensing milk on a glass slide and staining with methylene blue dye allowing somatic cell nuclei to be visualized and counted (Orlandini, 2011). Vaccines There are many reports that have investigated the use of vaccines for prevention of bovine mastitis (Pereira et al., 2011). Currently, few vaccines are commercially available for preventing mastitis. The obvious challenge to developing a broadly efficacious bovine mastitis vaccine is the large variety of bacteria known to cause mastitis (Watts, 1988). Another challenge to develop mastitis vaccines is inducing long term immunity when the target bacterium may change antigenically or with subtypes of bacteria specific to different regions and herds. Vaccine development has primarily targeted mastitis caused by Staphylococcus aureus and Escherichia coli (Pereira et al., 2011) likely due to the pathogenicity and high prevalence of cases (Makovec and Ruegg, 2003; Oliveira et al., 2013; USDA-APHIS, 2008). Pereira et al. published an extensive review of S. aureus vaccine studies and developed a scoring system to determine the efficacy of published methods. Their findings indicated that the use of recombinant proteins associated with S. aureus provided 50% protection in experimentally infected quarters. One of the reviewed studies, Carter and Kerr, 2003, performed this method using protein A, a virulence factor of S. aureus, encoded on a staphylococcal plasmid which was then transfected into cells that were injected into animals. Pereira et al. also reported 50% protection using inactivated vaccines known as bacterins or inactivated toxins known as toxoids. One such study (Leitner et al., 2003) achieved success

8 using a bacterin and toxoid from 3 different strains of S. aureus with different hemolysis patterns. Contagious and Environmental Mastitis Bovine mastitis is classified into two different types according to the origin of the organism and mode of exposure. Contagious mastitis is associated with several pathogens with the ability to be transmitted between cows and exist in low numbers in the environment. Environmental mastitis is caused by a variety of organisms found in the cow s environment originating from manure, soil, water or bedding. However, once a cow is infected, they will shed the organism into the environment or milking equipment with the ability to transmit the pathogen to another animal. Contagious Mastitis Staphylococcus aureus Perhaps the most significant mastitis pathogen, Staphylococcus aureus, has been isolated from individual cow and bulk tank milk samples in 43% of U.S. dairy operations in the top 17 milk-producing states according to a USDA-APHIS study in 2007. Mastitis caused by S. aureus may present as clinical, subclinical and chronic infections in cows (Srivastava et al., 2015). S. aureus is associated with diseases of several body systems in humans and has become a critical public health concern with its increasing resistance to antimicrobials. The Staphylococcus genus refers to a Gram-positive coccoid bacterium often arranged in groups or clusters. Colony morphology is large, white or yellow mucoid colonies on agar plates. Staphylococci are catalase positive and exhibit gamma or beta hemolysis although S. aureus is usually beta hemolytic. S. aureus is a common skin inhabitant of humans and the udder of cows. Most S. aureus infections affecting the skin are opportunistic,

9 requiring damage to the epidermis such as lacerations, abrasions and burns, to cause infection. S. aureus primarily affects dairy cows by intramammary infection and represents the most economically devastating mastitis pathogen (Oliver et al., 2004). Some species of staphylococci produce coagulase, which is an important surface protein that is commonly used to differentiate S. aureus from other Staphylococcus spp. although it is not the only coagulase-positive Staphylococcus species. Coagulase proteins produced by some staphylococci are considered a virulence factor although coagulase itself is not directly involved in pathogenesis. Coagulase induces clumping of the blood by forming a complex with fibrinogen and prothrombin, two blood proteins involved in the clotting cascade (Graber et al., 2009). Coagulase mediates adherence to mammalian cells, such as red blood cells, that is thought to disguise the bacterium as a host protein to prevent recognition by the immune system (Salyers and Whitt, 2002). Coagulase status is determined by rinsing pure culture into rabbit sera in a tube or on a slide. The formation of a gel or coagulation of the serum indicates the bacterium contains the coagulase enzyme (Boerlin et al., 2003). S. aureus has a variety of extracellular enzymes found in bovine mastitis isolates including staphylokinase, hyaluronidase, phosphatase, nuclease, lipase and catalase. These enzymes are thought to assist S. aureus in survival and spread in the host by destroying extracellular matrices or making milk components available as an energy source for the bacterium (Salyers and Whitt, 2002; Sutra and Poutrel, 1994). Exact mechanisms of their pathogenesis are not known. Streptococcus agalactiae Streptococcus agalactiae is a significant contagious pathogen in bovine mastitis with a high propensity to transmit from cow to cow, lacking the ability to thrive in the

10 environment. It causes clinical and subclinical mastitis in dairy cattle passing from cow to cow by the hands of milking staff, cleaning rags used on cows as well as milking machines (Hogan et al., 1999). When 530 U.S. dairy operations had milk from bulk tanks sampled, 2.6% cultured positive for S. agalactiae, a low percentage compared to S. aureus or E.coli which were both above 40% (USDA-APHIS, 2008). The Streptococcus genus represents nonmotile, gram-positive coccoid bacteria often arranged in chained patterns. These catalase negative bacteria tolerate oxygen but may prefer anaerobic conditions and perform fermentative metabolism. They exhibit alpha, beta, and gamma hemolysis in culture and present as small to medium sized colonies (Oliver et al., 2004). Little is understood about the virulence of S. agalactiae. One potential virulence factor was observed in human infections was found with the production of maternal antibody to its capsular polysaccharides suggesting the capsular polysaccharides may cause clinical symptoms (Salyers and Whitt, 2002). Bovine isolates contain the gene hylb encoding hyaluronidase, an enzyme capable of cleaving the extracellular matrix in tissues (Sukhnanand et al., 2005). Mycoplasma species Mycoplasma is a genus of bacteria associated with contagious bovine mastitis. The most prevalent Mycoplasma species is Mycoplasma bovis however, 11 additional species have been isolated from bovine mastitis (Oliver et al., 2004). Mycoplasma spp. is considered a fastidious organism that requires 10% CO2 in a 37 C incubator for 7 to 10 days on culture media (Hogan et al., 1999). Mycoplasma spp. lack a cell wall so are resistant to antibiotics that target the cell structure (Bürki et al., 2015).

11 Mycoplasma spp. also perpetrate respiratory infections, arthritis, otitis and genital infections in cows thus providing numerous potential reservoirs in a dairy operation (Bürki et al., 2015). Little is understood about the molecular mechanisms of pathogenicity of mycoplasmal species. Their small genomes provide a few processes that allow Mycoplasma spp. to adhere and survive in hosts. Adherence to host cells, heavily mediated by surface proteins, is critical for this organism as they lack biosynthetic pathways to acquire essential resources such as amino acids. M. bovis has been found to invade alveolar epithelial cells in the bovine mammary gland which may contribute to its ability to disseminate throughout the host (Bürki et al., 2015). Environmental mastitis Environmental Streptococcus spp. Mastitis caused by streptococci species are attributed to Streptococcus dysgalactiae and Streptococcus uberis. Other environmental streptococci known to cause mastitis are S. acidominus, S. canis, S. equi, and S. equinus. These pathogens have been isolated from the intestinal tract and manure of dairy cattle which may contaminate bedding (Oliver et al., 2004). In dairy operations, concentrations of streptococci in bedding may reach 10 6 CFU/g in wood shavings, recycled manure and pelleted corn (Todhunter et al., 1995). Pathogenicity of environmental streptococci is not well understood however several mechanisms allow streptococci to thwart host immunity and adhere to and invade host cells. S. dysgalactiae and S. uberis interact with host proteins fibronectin, fibrinogen, immunoglobulins, collagen and laminin which enables the bacteria to adhere to host cells. S. dysgalactiae contains hyaluronidase and fibrinolysin believed to allow the bacterium to disseminate in host tissue. The capsule of S. uberis allows it to avoid phagocytosis however,

12 similar capsule formation has not been observed in S. dysgalactiae. Both species have been found inside mammary secretory cells thought to involve host cell kinases and rearrangement of microfilaments (Calvinho et al., 1998). Coliforms Coliforms are gram-negative bacteria that ferment lactose to produce gas and are associated with the gastrointestinal system of animals. Coliforms are responsible for animal health and food safety concerns because they can be pathogenic and also reflect the hygienic practices of dairy operations and food processors. The presence of coliforms in any product suggests contamination with fecal material through manure, sewage or run-off. This material is likely to harbor bacteria responsible for foodborne-illness. Grade A milk may not contain greater than 10 coliform bacteria per ml of raw milk upon arrival at the processing plant according to the 2015 PMO (FDA, 2015). Coliform bacteria are abundant in the environment of dairy operations (Hogan et al., 1999). The most common coliforms that can cause mastitis are E. coli, Klebsiella pneumoniae or oxytoca, and to a lesser extent Citrobacter species and Enterobacter species (Makovec and Ruegg, 2003; Oliveira et al., 2013). Bacterial isolation from milk is often achieved with selective and differential media including MacConkey s agar which is selective for gram-negative organisms and differentiates them based on their lactose fermentation result. Growth appears at 18 hours when incubated at 37 C. Further identification is often performed with differential biochemical tests for gas production, motility and carbohydrate substrate utilization (Oliver et al., 2004). E. coli is the most common cause of coliform mastitis (Makovec and Ruegg, 2003; Oliveira et al., 2013) with the ability to induce mild to severe symptoms for a short duration.

13 The pathogen is often eliminated by the cow s immune system without treatment (Döpfer et al., 1999; Wenz et al., 2006). E. coli, like other gram negative organisms, possess an outer membrane composed of antigenic lipopolysaccharide (LPS) the component that induces the initial immune response and inflammation after infection (Salyers and Whitt, 2002). There are few consistent virulence factors among E. coli bovine mastitis isolates (Fernandes et al., 2011; Wenz et al., 2006) but they all have components that cause an immune reaction and allow better colonization of the organism (Shpigel et al., 2008). One of these is the ability to form biofilms on the mammary alveolar epithelial cells (Shpigel et al., 2008) thought to be mediated by type I fimbriae (Fernandes et al., 2011). Fernandes et al. further concluded that the majority of E. coli isolates analyzed from bovine mastitis were of a genotype associated with commensal E. coli found in the GI tract indicating cows are exposed when they come into contact with manure or soiled bedding. Klebsiella is another opportunistic bacterial species found in a dairy cow s environment. Klebsiella pneumoniae has become significant in human medicine in the ongoing battle with multidrug resistant bacteria in hospital environments (Diancourt et al., 2005). Klebsiella pneumoniae resists attack by PMNs due to a thick polysaccharide capsule production that surrounds the outer membrane of the bacterium (Kanevsky-Mullarky et al., 2014). This capsule also makes K. pneumoniae distinguishable on an agar plate due to the mucoid appearance of the colonies (Oliver et al., 2004). Enterococcus species Enterococcus species were originally thought to be members of the Streptococcus genus and innocuous to dairy cows. Enterococcus species has risen in clinical occurrence in hospital environments. Like streptococci, Enterococcus spp. are gram positive and catalase

14 negative. Where the two genera differ is their reservoirs in animals where enterococci are found in the intestinal tract and manure of cows (Oliver et al., 2004) at concentrations near that of E. coli (Salyers and Whitt, 2004). Viruses With nearly 30% of bovine milk samples submitted to diagnostic laboratories for bacteriological culture yielding a no growth result (Makovec and Ruegg, 2003; Oliveira et al., 2013) it is possible that a portion of these cases are attributable to viral infections. Most viral bovine pathogens do not produce an infection local to the mammary gland but are shed through milk due to a systemic viral infection. Viral pathogens associated with dairy cattle and isolated from milk include bovine herpes virus (BHV), foot and mouth disease (FMD), parainfluenza (PI), bovine leukemia virus, vaccinia, cow pox, vesicular stomatitis, bovine papilloma virus, bovine viral diarrhea virus, rinderpest virus and bovine enterovirus (Wellenberg et al., 2002). Clinical mammary symptoms present as classic inflammation such as swelling and tenderness of the udder, fever and failure to thrive. Subclinical signs include an increased SCC and decreased milk production. Viruses BHV1, FMD, PI3 have induced clinical mastitis after experimental introduction through the intramammary route and BHV4 has induced subclinical mastitis. However, natural induction through the mammary gland is not common due to their susceptibility to the environment (Wellenberg et al., 2002). Diagnosis of bovine mastitis Dairy practitioners and herd managers use a variety of tools to diagnose mastitis in an animal before the next step of pathogen detection. Most often, somatic cell count (SCC), which can be determined by several methods, is used to detect mastitis. Other methods such

15 as metal conductivity, visual inspection and animal records assist clinicians in identifying mastitis cases. PCR/nucleic acid tests for the detection of bovine mastitis The characterization of polymerase chain reaction (PCR) in 1985 by Kerry Mullis incited a substantial change to virtually all areas of biological science. PCR allowed scientists to choose a segment of deoxyribonucleic acid (DNA) and exponentially increase the amount of DNA in a sample for further analysis. The first nucleic acid assays for use with bovine milk were developed to detect pathogens that are difficult to cultivate in the laboratory. Various PCR assays were used to detect Mycobacterium avium ssp. paratuberculosis (Johne s disease) (Over et al., 2011), Mycobacterium bovis (Antognoli et al., 2001; Sreevatsan et al., 2000), Brucella spp. (Romero and Lopez-Goñi 1999; Sreevatsan et al., 2000), Coxiella burnetii (Muramatsu et al., 1997), Cryptosporodium spp. (Laberge et al., 1996), and Mycoplasma spp. (Baird et al., 1999). Before nucleic acid assays can be performed the genetic material must be extracted and purified from the sample. Protocols developed for bacterial DNA extraction directly from clinical samples must reduce or eliminated inhibiting substances naturally occurring in the sample that can reduce the sensitivity of the assay. Milk contains fat, carbohydrates, protein and minerals that impede the isolation of bacterial DNA. Calcium ions in milk can affect DNA replication by interfering with the magnesium cofactor of polymerase. Milk proteins impede DNA polymerase by acting as a physical barrier by sequestering the target DNA and primers. Bacterial cellular debris and polysaccharides have shown similar ability to physically block DNA polymerase by the target DNA and primers (Wilson, 1997). Raw milk

16 samples are often combined with enrichment media to dilute these factors (Gillespie and Oliver, 2005; Rossen et al., 1992). For bovine mastitis diagnostics, PCR was originally used for the detection of Mycoplasma spp., an optimal candidate for PCR diagnostics due to its fastidious nature in culture. In 1999, Baird et al. developed a nested PCR assay to detect Mycoplasma spp. DNA present in bulk tank milk, individual cow milk and enriched liquid cultures with success. In a 2001 study by Riffon et. al. a multiplex PCR assay able to detect several mastitis pathogens was evaluated. S. aureus, S. agalactiae, S. dysgalactiae, S. uberis, S. parauberis and E. coli cultures were inoculated into UHT (ultra-high temperature) pasteurized milk where samples were exposed to methods including a pre-enzymatic lysis and methods without a preenzymatic lysis step. Two different primer sequences were used for each pathogen target representing the16s and 23s rrna genes to ensure specificity between related species. High specificity was achieved in this analysis with all targets properly amplified and detected and non-targets presenting no amplification. Analytical sensitivity of the assay was 3.12x10 2 CFU/mL of milk with the pre-pcr enzymatic lysis and 5x10 3 CFU/mL of milk from the process that did not include a pre-pcr enzymatic lysis step. The authors chose to investigate the effect on sensitivity without a pre-pcr lysis step to reduce the costs of reagents. It is evident that without a pre-pcr lysis step some sensitivity is sacrificed and will have to be considered with the assay s use. Another multiplex PCR analysis was used for the detection of S. agalactiae, S. uberis, S. dysgalactiae, and S. aureus in bulk tank milk samples in Australia. Phuektes et al. sampled bulk tanks every 10 days for 5 collections from 42 farms for 176 samples comparing PCR to bulk tank somatic cell count (BTMSCC) and total bacteria count (TBC). Although the study

17 did not directly compare culture and PCR for the detection of S. agalactiae, S. uberis, S. dysgalactiae, and S. aureus, results indicated that PCR was able to detect the four organisms from bulk tank milk according to their methods. Real-time multiplex PCR was first studied for use with bovine milk for mastitis pathogens in 2005 by Gillespie and Oliver. Species-specific primer/probe sets were designed based on separate studies validating individual real-time PCR assays for the detection of S. agalactiae, S. uberis and S. aureus. The study compared 2 commercial methods and one additional method of extracting bacterial DNA from milk samples. The most reproducible results came from the use of the enzyme pronase which acts on the casein protein, dismantling the micelle that forms with calcium and phosphorus ions, which can interfere with the PCR reaction. Analytical specificity was 100% performed with 47 non-target ATCC strains of 31 target bacterial species. Limit of detection (LOD) or analytical sensitivity was determined using UHT pasteurized milk inoculated with the three organisms with subsequent serial dilutions. The LOD of S. aureus was 10 3 CFU/mL and the LOD for S. agalactiae and S. uberis was 10 2 CFU/mL. All 3 organisms achieved an LOD of 10 0 CFU/mL with an overnight enrichment step. The assay was then tested against 192 mastitis quarter milk samples with conventional culture results indicating an analytical sensitivity of 95.5% and specificity of 99.6% for target and non-target pathogens. Increased sensitivity was observed when 20 S. aureus and S. uberis positive milk samples were enriched with culture broth and incubated overnight and compared to non-enrichment PCR results. It was proposed that the improvement was due to dilution of the inhibitory substances found in milk and the increased growth due to the addition of nutrient broth and incubation time. The potential concern with enriching milk samples is the high likelihood of milk samples being contaminated from the

18 environment or the udder skin. Enrichment allows for the propagation of environmental contaminants that are in low numbers in the original sample and when tested with PCR could deliver a false positive result prolonging testing time. PathoProof Mastitis PCR PathoProof Mastitis PCR was the first commercial multiplex, real-time PCR assay capable of simultaneously detecting eleven bovine mastitis pathogens and one resistance gene. The test was developed by Finnzymes Oy of Espoo, Finland and is currently manufactured by ThermoFisher Scientific Inc. of Waltham, Massachusetts. The assay is contained in a kit providing all the reagents needed to extract bacterial DNA from bovine milk and perform real-time PCR. The assay is offered in several forms designated the major- 3, complete-12 and complete-16, each identifying significant bovine mastitis pathogens. All kits are compatible with different extraction equipment and thermocycler systems including the KingFisher 96 deepwell automatic extraction using magnetic particle processing and the Applied Biosystems 7500 Fast thermocycler (both manufactured by ThermoFisher Scientific Inc., Waltham, MA). The complete-12 assay is performed in 4 separate reactions containing 3 different primer/probe sets to detect their respective targets. Included in each reaction is an internal amplification control (IAC), which is a fragment of DNA that in the absence of PCR inhibition should amplify during the reaction and act as a positive control. The 4 separate reactions are as follows: primer/probe set 1 detects Staphylococcus aureus, Enterococcus species (including E. faecalis and E. faecium), Corynebacterium bovis and an IAC primer/probe; set 2 detects Staphylococcus blaz, the gene encoding beta lactamase responsible for resistance to beta-lactam antibiotics, Escherichia coli, Streptococcus dysgalactiae and an IAC; primer/probe set 3 detects Staphylococcus species (including

19 coagulase-negative Staphylococcus spp.), Streptococcus agalactiae, Streptococcus uberis and an IAC; primer/probe set 4 detects Klebsiella species (including K. oxytoca and K. pneumoniae), Serratia marcescens, Trueperella pyogenes and/or Peptoniphilus indolicus and an IAC. PathoProof assays include the Norden Lab Mastitis Studio software that provides interpretation of the data collected by thermocycler systems. The Applied Biosystems 7500 Fast system collects data using sequence detection software (SDS) which can be uploaded into the Norden software. The data is reported in a standard format provided by Norden that includes a list of targets that were determined positive in the assay and their corresponding Ct values. The process is simple and produces results in a report suitable for a client. Currently, there are publications describing the use of Norden software for analysis with the PathoProof assay (Keane et al, 2013; Cervinkova et al, 2013) but not including in-depth discussion of its use and result interpretation leaving question as to its accuracy and utility. The first published investigation of PathoProof PCR was in 2007 by Pitkälä et al. when it was used in a comparison of different assays to detect beta-lactamase-producing Staphylococcus species. One hundred and seventy-five Staphylococcus spp. isolates were used in the comparison including S. aureus, S. intermedius, and coagulase-negative species including S. epidermidis, S. chromogenes, S. cohnii, S. xylosus, S. hyicus, S. haemolyticus, S. warneri, S. saprophyticus and S. simulans all originating from cases of clinical bovine and canine mastitis samples. The assays evaluated in this study were compared to a PCR reference method for the detection of the beta-lactamase gene sequence. All methods, except for PathoProof PCR, compared in this analysis were designed to detect the beta-lactamase enzyme while PathoProof PCR was the only assay included to detect the blaz gene.

20 Compared to the reference method, PathoProof PCR detected all isolates with the betalactamase gene and did not detect the isolates that were determined negative by the reference method. PathoProof PCR was the most successful method in detecting beta-lactamase producing Staphylococcus spp. of all the methods compared in the study. The authors recommended PathoProof for routine beta-lactamase producing Staphylococcus testing however given that the assay is multiplexed and developed to extract bacterial DNA from bovine milk samples making it was impractical for the detection of single pathogens isolated from pure culture (Pitkälä et al., 2007). An extensive validation study of PathoProof PCR was published in 2009 by Koskinen et al. with the title, Analytical specificity and sensitivity of a real-time polymerase chain reaction assay for identification of bovine mastitis pathogens. Koskinen et al. collected 643 culture isolates originating from bovine, human and companion animal mastitic milk samples from diagnostic laboratories dispersed over a wide geographical range. Targets included 525 isolates including 72 blaz gene-positive staphylococci isolates used and validated by Pitkälä et al., 2007. One-hundred and eighteen isolates were non-target species. The original culture results obtained by the respective laboratories were compared to PathoProof PCR results that became the basis of Koskinen et al. specificity and sensitivity analysis. In cases where target isolates did not match their PathoProof PCR result, 16s rrna sequencing and comparison was completed. PathoProof PCR was successful in properly identifying 634 of the 643 isolates including all isolates originating from bovine mastitis with no false negative results. Nine isolates were found falsely positive by PathoProof PCR as confirmed by 16s rrna sequencing including Streptococcus pyogenes, Streptococcus sanguis, Streptococcus

21 salivarius identified as S. uberis, and 1 Shigella spp. isolate identified as E. coli. The authors of the study reported a 100% analytical specificity for all PathoProof targets except for S. uberis at 99% and E. coli at 99.5% and a 100% analytical sensitivity for all targets using false positive and negative equations. It should be noted that this study did not apply a consistent, standard method to identify all culture isolates. The objectives in this study were to define the analytical sensitivity and specificity of the PathoProof assay however analytical sensitivity is defined as the lowest concentration that is detectable by the assay, expressed in a numerical value (Saah and Hoover, 1997). Such figure was not described in this study despite its title. What was performed in this study more closely resembled a diagnostic sensitivity analysis by testing closely-related, non-target species however diagnostic sensitivity for PathoProof should be determined using clinical milk samples according to the definition of diagnostic sensitivity (Saah and Hoover, 1997). Analytical sensitivity is important to diagnosticians as some samples may have low concentrations of target. A threshold concentration of bacteria in a sample that signifies bovine mastitis has not been established in any literature therefore an established analytical sensitivity of PathoProof PCR may not directly translate into a diagnosis. However, analytical sensitivity is still valuable when evaluating an assay and that has yet to be determined for the PathoProof PCR. PathoProof PCR research continued with Real-time polymerase chain reaction-based identification of bacteria in milk samples from bovine clinical mastitis with no growth in conventional culturing published in 2009. The objectives in this study were to address PathoProof PCR as a potential solution to the frequent culture-negative results in cows with clinical mastitis. Using 79 milk samples that were negative or no-growth in culture, the

22 authors used PathoProof PCR to detect bacteria in 43% of the milk samples tested. The Ct values obtained for the bacterial targets ranged from 22.2 to 36.7 with an average of 32.3. Detected targets included S. uberis (10), S. dysgalactiae (2), S. aureus (3), T. pyogenes (1), E. faecalis/faecium (1), E. coli (1), Staphylococcus species (9) and C. bovis (5) with some samples positive for 2 targets. The average Ct was relatively high indicating low concentration of target which is consistent with the negative culture results that are attributed to low concentrations of bacteria in a sample. Of the positive targets discretion may be needed for culture-negative samples that are positive for Staphylococcus spp. and C. bovis as they are 2 common skin inhabitants making them common contaminants in milk collections. A separate Ct threshold of significance for these 2 organisms may improve interpretation. This study also investigated the quantitative abilities of PathoProof PCR by creating a standard curve to find the concentration of copy numbers in a sample. The authors used the kit s amplification standard, DNA concentration, amplicon length and mass to calcluate the amount of genome copies per ml of milk. The lowest Ct value among all of the samples was 22.2 for Staphylococcus spp. which corresponded to 1.7x10 3 copies in the original milk sample. The results of this study are helpful for interpretation of Ct values for diagnostic purposes however similar work with empirical methods to determine the LOD is necessary to understand PathoProof PCR abilities. Another study using bovine milk samples was published in 2010 by Koskinen et al. comparing culture to PathoProof PCR. The authors collected 780 quarter milk samples from clinical mastitis cases as well as 220 samples from cows without clinical indications of mastitis. Culture methods consisted of selective and differential media to isolate and identify